@gajae-code/ai 0.1.1 → 0.1.3

package/CHANGELOG.md

@@ -2,6 +2,18 @@
 
 ## [Unreleased]
 
+## [0.1.3] - 2026-05-28
+
+### Changed
+
+- Released the current dev branch fixes with refreshed 0.1.3 package metadata.
+
+## [0.1.2] - 2026-05-28
+
+### Changed
+
+- Updated package metadata for the Gajae Code npm publication.
+
 ## [0.1.1] - 2026-05-28
 ### Breaking Changes
 

package/dist/types/auth-storage.d.ts

@@ -390,6 +390,8 @@ export declare class AuthStorage {
      * Reload credentials from storage.
      */
     reload(): Promise<void>;
+    /** Returns the credential type selected for a provider/session, if one has been recorded. */
+    getSessionCredentialType(provider: string, sessionId?: string): AuthCredential["type"] | undefined;
     /**
      * Get credential for a provider (first entry if multiple).
      */

package/dist/types/types.d.ts

@@ -233,6 +233,11 @@ export interface StreamOptions {
      * channel) silently ignore the override.
      */
     fetch?: FetchImpl;
+    /**
+     * Authentication credential type selected for this request.
+     * Providers use this only when endpoint routing differs between API-key and OAuth credentials.
+     */
+    authCredentialType?: "api_key" | "oauth";
     /** Cursor exec/MCP tool handlers (cursor-agent only). */
     execHandlers?: CursorExecHandlers;
 }

package/package.json

@@ -1,10 +1,10 @@
 {
 	"type": "module",
 	"name": "@gajae-code/ai",
-	"version": "0.1.1",
+	"version": "0.1.3",
 	"description": "Unified LLM API with automatic model discovery and provider configuration",
-	"homepage": "https://gajae-code.dev",
-	"author": "Can Boluk",
+	"homepage": "https://gaebal-gajae.dev",
+	"author": "Yeachan-Heo",
 	"contributors": [
 		"Mario Zechner"
 	],
@@ -43,7 +43,7 @@
 	"dependencies": {
 		"@anthropic-ai/sdk": "^0.94.0",
 		"@bufbuild/protobuf": "^2.12.0",
-		"@gajae-code/utils": "0.1.1",
+		"@gajae-code/utils": "0.1.3",
 		"openai": "^6.36.0",
 		"partial-json": "^0.1.7",
 		"zod": "4.4.3"

package/src/auth-storage.ts

@@ -1046,6 +1046,11 @@ export class AuthStorage {
 		return this.#sessionLastCredential.get(provider)?.get(sessionId);
 	}
 
+	/** Returns the credential type selected for a provider/session, if one has been recorded. */
+	getSessionCredentialType(provider: string, sessionId?: string): AuthCredential["type"] | undefined {
+		return this.#getSessionCredential(provider, sessionId)?.type;
+	}
+
 	/** Clears the last credential used by a session for a provider. */
 	#clearSessionCredential(provider: string, sessionId: string | undefined): void {
 		if (!sessionId) return;

package/src/models.json

@@ -8452,7 +8452,7 @@
 			"cost": {
 				"input": 1,
 				"output": 3,
-				"cacheRead": 0,
+				"cacheRead": 0.5,
 				"cacheWrite": 0
 			},
 			"contextWindow": 262144,
@@ -8471,7 +8471,7 @@
 			"cost": {
 				"input": 0.15,
 				"output": 0.6,
-				"cacheRead": 0,
+				"cacheRead": 0.075,
 				"cacheWrite": 0
 			},
 			"contextWindow": 131072,
@@ -8495,7 +8495,7 @@
 			"cost": {
 				"input": 0.075,
 				"output": 0.3,
-				"cacheRead": 0,
+				"cacheRead": 0.0375,
 				"cacheWrite": 0
 			},
 			"contextWindow": 131072,
@@ -12861,6 +12861,25 @@
 				"maxLevel": "xhigh"
 			}
 		},
+		"moonshotai/kimi-k2.6:free": {
+			"id": "moonshotai/kimi-k2.6:free",
+			"name": "MoonshotAI: Kimi K2.6 (free)",
+			"api": "openai-completions",
+			"provider": "kilo",
+			"baseUrl": "https://api.kilo.ai/api/gateway",
+			"reasoning": false,
+			"input": [
+				"text"
+			],
+			"cost": {
+				"input": 0,
+				"output": 0,
+				"cacheRead": 0,
+				"cacheWrite": 0
+			},
+			"contextWindow": 222222,
+			"maxTokens": 8888
+		},
 		"morph-warp-grep-v2": {
 			"id": "morph-warp-grep-v2",
 			"name": "Morph: WarpGrep V2",
@@ -51146,7 +51165,7 @@
 			"name": "OpenAI code Mini",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text"
@@ -51170,7 +51189,7 @@
 			"name": "GPT-4",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": false,
 			"input": [
 				"text"
@@ -51189,7 +51208,7 @@
 			"name": "GPT-4 Turbo",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": false,
 			"input": [
 				"text",
@@ -51209,7 +51228,7 @@
 			"name": "GPT-4.1",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": false,
 			"input": [
 				"text",
@@ -51229,7 +51248,7 @@
 			"name": "GPT-4.1 mini",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": false,
 			"input": [
 				"text",
@@ -51249,7 +51268,7 @@
 			"name": "GPT-4.1 nano",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": false,
 			"input": [
 				"text",
@@ -51269,7 +51288,7 @@
 			"name": "GPT-4o",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": false,
 			"input": [
 				"text",
@@ -51289,7 +51308,7 @@
 			"name": "GPT-4o (2024-05-13)",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": false,
 			"input": [
 				"text",
@@ -51309,7 +51328,7 @@
 			"name": "GPT-4o (2024-08-06)",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": false,
 			"input": [
 				"text",
@@ -51329,7 +51348,7 @@
 			"name": "GPT-4o (2024-11-20)",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": false,
 			"input": [
 				"text",
@@ -51349,7 +51368,7 @@
 			"name": "GPT-4o mini",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": false,
 			"input": [
 				"text",
@@ -51369,7 +51388,7 @@
 			"name": "GPT-5",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51394,7 +51413,7 @@
 			"id": "gpt-5-chat-latest",
 			"name": "GPT-5 Chat Latest",
 			"api": "openai-responses",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"provider": "openai",
 			"reasoning": false,
 			"input": [
@@ -51416,7 +51435,7 @@
 			"name": "GPT-5-OpenAI code",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51442,7 +51461,7 @@
 			"name": "GPT-5 Mini",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51468,7 +51487,7 @@
 			"name": "GPT-5 Nano",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51494,7 +51513,7 @@
 			"name": "GPT-5 Pro",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51520,7 +51539,7 @@
 			"name": "GPT-5.1",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51546,7 +51565,7 @@
 			"name": "GPT-5.1 Chat",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51572,7 +51591,7 @@
 			"name": "GPT-5.1 OpenAI code",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51598,7 +51617,7 @@
 			"name": "GPT-5.1 OpenAI code Max",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51624,7 +51643,7 @@
 			"name": "GPT-5.1 OpenAI code mini",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51650,7 +51669,7 @@
 			"name": "GPT-5.2",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51676,7 +51695,7 @@
 			"name": "GPT-5.2 Chat",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51702,7 +51721,7 @@
 			"name": "GPT-5.2 OpenAI code",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51728,7 +51747,7 @@
 			"name": "GPT-5.2 Pro",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51754,7 +51773,7 @@
 			"name": "GPT-5.3 Chat (latest)",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": false,
 			"input": [
 				"text",
@@ -51775,7 +51794,7 @@
 			"name": "GPT-5.3 OpenAI code",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51801,7 +51820,7 @@
 			"name": "GPT-5.3 OpenAI code Spark",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51828,7 +51847,7 @@
 			"name": "GPT-5.4",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51854,7 +51873,7 @@
 			"name": "GPT-5.4 mini",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51880,7 +51899,7 @@
 			"name": "GPT-5.4 nano",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51906,7 +51925,7 @@
 			"name": "GPT-5.4 Pro",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51932,7 +51951,7 @@
 			"name": "GPT-5.5",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51959,7 +51978,7 @@
 			"name": "GPT-5.5 Pro",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -51986,7 +52005,7 @@
 			"name": "o1",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -52011,7 +52030,7 @@
 			"name": "o1-pro",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -52036,7 +52055,7 @@
 			"name": "o3",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -52061,7 +52080,7 @@
 			"name": "o3-deep-research",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -52086,7 +52105,7 @@
 			"name": "o3-mini",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text"
@@ -52110,7 +52129,7 @@
 			"name": "o3-pro",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -52135,7 +52154,7 @@
 			"name": "o4-mini",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -52160,7 +52179,7 @@
 			"name": "o4-mini-deep-research",
 			"api": "openai-responses",
 			"provider": "openai",
-			"baseUrl": "https://api.openai.com/v1",
+			"baseUrl": "",
 			"reasoning": true,
 			"input": [
 				"text",
@@ -52973,9 +52992,9 @@
 				"image"
 			],
 			"cost": {
-				"input": 0.4,
-				"output": 2,
-				"cacheRead": 0.08,
+				"input": 0.14,
+				"output": 0.28,
+				"cacheRead": 0.0028,
 				"cacheWrite": 0
 			},
 			"contextWindow": 1000000,
@@ -52997,9 +53016,9 @@
 				"text"
 			],
 			"cost": {
-				"input": 1,
-				"output": 3,
-				"cacheRead": 0.2,
+				"input": 1.74,
+				"output": 3.48,
+				"cacheRead": 0.0145,
 				"cacheWrite": 0
 			},
 			"contextWindow": 1048576,
@@ -54267,6 +54286,31 @@
 				"maxLevel": "xhigh"
 			}
 		},
+		"mimo-v2.5-free": {
+			"id": "mimo-v2.5-free",
+			"name": "MiMo V2.5 Free",
+			"api": "openai-completions",
+			"provider": "opencode-zen",
+			"baseUrl": "https://opencode.ai/zen/v1",
+			"reasoning": true,
+			"input": [
+				"text",
+				"image"
+			],
+			"cost": {
+				"input": 0,
+				"output": 0,
+				"cacheRead": 0,
+				"cacheWrite": 0
+			},
+			"contextWindow": 1000000,
+			"maxTokens": 128000,
+			"thinking": {
+				"mode": "effort",
+				"minLevel": "minimal",
+				"maxLevel": "xhigh"
+			}
+		},
 		"minimax-m2.1": {
 			"id": "minimax-m2.1",
 			"name": "MiniMax M2.1",
@@ -57044,7 +57088,7 @@
 				"cacheRead": 0,
 				"cacheWrite": 0
 			},
-			"contextWindow": 204800,
+			"contextWindow": 262144,
 			"maxTokens": 8192,
 			"compat": {
 				"supportsToolChoice": false
@@ -57726,6 +57770,31 @@
 				"maxLevel": "high"
 			}
 		},
+		"moonshotai/kimi-k2.6:free": {
+			"id": "moonshotai/kimi-k2.6:free",
+			"name": "MoonshotAI: Kimi K2.6 (free)",
+			"api": "openai-completions",
+			"provider": "openrouter",
+			"baseUrl": "https://openrouter.ai/api/v1",
+			"reasoning": true,
+			"input": [
+				"text",
+				"image"
+			],
+			"cost": {
+				"input": 0,
+				"output": 0,
+				"cacheRead": 0,
+				"cacheWrite": 0
+			},
+			"contextWindow": 262144,
+			"maxTokens": 8888,
+			"thinking": {
+				"mode": "effort",
+				"minLevel": "minimal",
+				"maxLevel": "high"
+			}
+		},
 		"nex-agi/deepseek-v3.1-nex-n1": {
 			"id": "nex-agi/deepseek-v3.1-nex-n1",
 			"name": "Nex AGI: DeepSeek V3.1 Nex N1",
@@ -58616,11 +58685,11 @@
 			"cost": {
 				"input": 1.25,
 				"output": 10,
-				"cacheRead": 0.125,
+				"cacheRead": 0.13,
 				"cacheWrite": 0
 			},
 			"contextWindow": 128000,
-			"maxTokens": 16384
+			"maxTokens": 32000
 		},
 		"openai/gpt-5.1-codex": {
 			"id": "openai/gpt-5.1-codex",
@@ -58636,7 +58705,7 @@
 			"cost": {
 				"input": 1.25,
 				"output": 10,
-				"cacheRead": 0.125,
+				"cacheRead": 0.13,
 				"cacheWrite": 0
 			},
 			"contextWindow": 272000,
@@ -58686,11 +58755,11 @@
 			"cost": {
 				"input": 0.25,
 				"output": 2,
-				"cacheRead": 0.03,
+				"cacheRead": 0.024999999999999998,
 				"cacheWrite": 0
 			},
 			"contextWindow": 272000,
-			"maxTokens": 128000,
+			"maxTokens": 100000,
 			"thinking": {
 				"mode": "effort",
 				"minLevel": "medium",
@@ -58740,7 +58809,7 @@
 				"cacheWrite": 0
 			},
 			"contextWindow": 128000,
-			"maxTokens": 32000
+			"maxTokens": 16384
 		},
 		"openai/gpt-5.2-codex": {
 			"id": "openai/gpt-5.2-codex",
@@ -61027,13 +61096,13 @@
 				"text"
 			],
 			"cost": {
-				"input": 0.06599999999999999,
-				"output": 0.26,
-				"cacheRead": 0.029,
+				"input": 0.063,
+				"output": 0.21,
+				"cacheRead": 0.020999999999999998,
 				"cacheWrite": 0
 			},
 			"contextWindow": 262144,
-			"maxTokens": 262144,
+			"maxTokens": 64000,
 			"thinking": {
 				"mode": "effort",
 				"minLevel": "minimal",

package/src/provider-models/openai-compat.ts

@@ -1,3 +1,4 @@
+import { $env } from "@gajae-code/utils";
 import type { ModelManagerOptions } from "../model-manager";
 import { Effort } from "../model-thinking";
 import { getBundledModels } from "../models";
@@ -14,6 +15,7 @@ import { createBundledReferenceMap, createReferenceResolver } from "./bundled-re
 
 const MODELS_DEV_URL = "https://models.dev/api.json";
 const ANTHROPIC_BASE_URL = "https://api.anthropic.com/v1";
+const OPENAI_DEFAULT_BASE_URL = "https://api.openai.com/v1";
 const ANTHROPIC_OAUTH_BETA =
 	"claude-code-20250219,oauth-2025-04-20,interleaved-thinking-2025-05-14,context-management-2025-06-27,prompt-caching-scope-2026-01-05";
 
@@ -494,7 +496,7 @@ export interface OpenAIModelManagerConfig {
 
 export function openaiModelManagerOptions(config?: OpenAIModelManagerConfig): ModelManagerOptions<"openai-responses"> {
 	const apiKey = config?.apiKey;
-	const baseUrl = config?.baseUrl ?? "https://api.openai.com/v1";
+	const baseUrl = config?.baseUrl?.trim() || $env.OPENAI_BASE_URL?.trim() || OPENAI_DEFAULT_BASE_URL;
 	const references = createBundledReferenceMap<"openai-responses">("openai");
 	return {
 		providerId: "openai",
@@ -2105,7 +2107,7 @@ const MODELS_DEV_PROVIDER_DESCRIPTORS_CORE: readonly ModelsDevProviderDescriptor
 		"https://generativelanguage.googleapis.com/v1beta",
 	),
 	// --- OpenAI ---
-	simpleModelsDevDescriptor("openai", "openai", "openai-responses", "https://api.openai.com/v1"),
+	simpleModelsDevDescriptor("openai", "openai", "openai-responses", ""),
 	// --- Groq ---
 	openAiCompletionsDescriptor("groq", "groq", "https://api.groq.com/openai/v1"),
 	// --- Cerebras ---

package/src/providers/openai-completions.ts

@@ -70,6 +70,21 @@ import { createInitialResponsesAssistantMessage } from "./openai-responses-share
 import { transformMessages } from "./transform-messages";
 import { joinTextWithImagePlaceholder, NON_VISION_IMAGE_PLACEHOLDER } from "./vision-guard";
 
+const OPENAI_DEFAULT_BASE_URL = "https://api.openai.com/v1";
+
+function resolveOpenAIProviderBaseUrl(
+	baseUrl: string | undefined,
+	authCredentialType: "api_key" | "oauth" | undefined,
+): string {
+	if (authCredentialType === "oauth") return OPENAI_DEFAULT_BASE_URL;
+	const envBaseUrl = $env.OPENAI_BASE_URL?.trim();
+	const configuredBaseUrl = baseUrl?.trim();
+	if (envBaseUrl && (!configuredBaseUrl || configuredBaseUrl.toLowerCase().includes("api.openai.com"))) {
+		return envBaseUrl;
+	}
+	return configuredBaseUrl || envBaseUrl || OPENAI_DEFAULT_BASE_URL;
+}
+
 /**
  * Normalize tool call ID for Mistral.
  * Mistral requires tool IDs to be exactly 9 alphanumeric characters (a-z, A-Z, 0-9).
@@ -413,6 +428,7 @@ export const streamOpenAICompletions: StreamFunction<"openai-completions"> = (
 				options?.onSseEvent,
 				options?.fetch,
 				options?.streamFirstEventTimeoutMs,
+				options?.authCredentialType,
 			);
 			const premiumRequestsTotal = copilotPremiumRequests;
 			getCapturedErrorResponse = captureErrorResponse;
@@ -888,6 +904,7 @@ async function createClient(
 	onSseEvent?: OpenAICompletionsOptions["onSseEvent"],
 	fetchOverride?: FetchImpl,
 	streamFirstEventTimeoutOverride?: number,
+	authCredentialType?: OpenAICompletionsOptions["authCredentialType"],
 ): Promise<{
 	client: OpenAI;
 	copilotPremiumRequests: number | undefined;
@@ -916,7 +933,7 @@ async function createClient(
 		// SDK UA so traffic is identifiable in upstream provider logs.
 		// https://openrouter.ai/docs/app-attribution
 		headers["User-Agent"] = `Gajae-Code/${packageJson.version}`;
-		headers["HTTP-Referer"] = "https://gajae-code.dev/";
+		headers["HTTP-Referer"] = "https://gaebal-gajae.dev/";
 		headers["X-OpenRouter-Title"] = "Gajae-Code";
 		headers["X-OpenRouter-Categories"] = "cli-agent";
 		// Always-on response caching: identical requests return cached responses for free.
@@ -931,7 +948,8 @@ async function createClient(
 	}
 	let copilotPremiumRequests: number | undefined;
 
-	let baseUrl = model.baseUrl;
+	let baseUrl =
+		model.provider === "openai" ? resolveOpenAIProviderBaseUrl(model.baseUrl, authCredentialType) : model.baseUrl;
 	if (model.provider === "github-copilot") {
 		apiKey = parseGitHubCopilotApiKey(rawApiKey).accessToken;
 		const hasImages = hasCopilotVisionInput(context.messages);

package/src/providers/openai-responses.ts

@@ -102,6 +102,20 @@ export interface OpenAIResponsesOptions extends StreamOptions {
 const OPENAI_RESPONSES_PROVIDER_SESSION_STATE_PREFIX = "openai-responses:";
 const OPENAI_RESPONSES_FIRST_EVENT_TIMEOUT_MESSAGE =
 	"OpenAI responses stream timed out while waiting for the first event";
+const OPENAI_DEFAULT_BASE_URL = "https://api.openai.com/v1";
+
+function resolveOpenAIProviderBaseUrl(
+	baseUrl: string | undefined,
+	authCredentialType: "api_key" | "oauth" | undefined,
+): string {
+	if (authCredentialType === "oauth") return OPENAI_DEFAULT_BASE_URL;
+	const envBaseUrl = $env.OPENAI_BASE_URL?.trim();
+	const configuredBaseUrl = baseUrl?.trim();
+	if (envBaseUrl && (!configuredBaseUrl || configuredBaseUrl.toLowerCase().includes("api.openai.com"))) {
+		return envBaseUrl;
+	}
+	return configuredBaseUrl || envBaseUrl || OPENAI_DEFAULT_BASE_URL;
+}
 
 const OPENAI_RESPONSES_PROGRESS_EVENT_TYPES = new Set([
 	"response.created",
@@ -213,6 +227,7 @@ export const streamOpenAIResponses: StreamFunction<"openai-responses"> = (
 				cacheSessionId,
 				options?.onSseEvent,
 				options?.fetch,
+				options?.authCredentialType,
 			);
 			const premiumRequestsTotal = copilotPremiumRequests;
 			const providerSessionState = getOpenAIResponsesProviderSessionState(model, options?.providerSessionState);
@@ -224,7 +239,7 @@ export const streamOpenAIResponses: StreamFunction<"openai-responses"> = (
 				api: output.api,
 				model: model.id,
 				method: "POST",
-				url: `${baseUrl ?? "https://api.openai.com/v1"}/responses`,
+				url: `${baseUrl}/responses`,
 				body: params,
 			};
 			const openaiStream = await callWithCopilotModelRetry(
@@ -313,6 +328,7 @@ function createClient(
 	sessionId?: string,
 	onSseEvent?: OpenAIResponsesOptions["onSseEvent"],
 	fetchOverride?: FetchImpl,
+	authCredentialType?: OpenAIResponsesOptions["authCredentialType"],
 ): {
 	client: OpenAI;
 	copilotPremiumRequests: number | undefined;
@@ -331,7 +347,8 @@ function createClient(
 	const headers = { ...(model.headers ?? {}), ...(extraHeaders ?? {}) };
 	let copilotPremiumRequests: number | undefined;
 
-	let baseUrl = model.baseUrl;
+	let baseUrl =
+		model.provider === "openai" ? resolveOpenAIProviderBaseUrl(model.baseUrl, authCredentialType) : model.baseUrl;
 	if (model.provider === "github-copilot") {
 		apiKey = parseGitHubCopilotApiKey(rawApiKey).accessToken;
 		const hasImages = hasCopilotVisionInput(context.messages);

package/src/types.ts

@@ -364,6 +364,11 @@ export interface StreamOptions {
 	 * channel) silently ignore the override.
 	 */
 	fetch?: FetchImpl;
+	/**
+	 * Authentication credential type selected for this request.
+	 * Providers use this only when endpoint routing differs between API-key and OAuth credentials.
+	 */
+	authCredentialType?: "api_key" | "oauth";
 	/** Cursor exec/MCP tool handlers (cursor-agent only). */
 	execHandlers?: CursorExecHandlers;
 }